A Next Generation Publication Platform for Open Science

The scientific enterprise is a highly collaborative effort that is critically dependent on the accurate, clear, and truthful communication of technical results.  There is a growing recognition that “Open Science” practices, which involve sharing data, code, models, physical samples, and knowledge, is important for accelerating scientific progress and should be integrated in the publication process. In this SBIR project we propose to develop a next-generation, open-source publication platform that assists researchers, editors, and reviewers with a focus on supporting open science practices that extend and improve technical communication.  Our work will build on our substantial experience in both academic publishing and artificial intelligence (AI).  Today, the practice of open science can require substantial additional effort when preparing and vetting scientific papers, but the significant recent advances in AI make this an ideal time to apply Large Language Models and related methods to streamline the publication process and make it much easier to meet the requirements of open science. Our work in this project will include 1) conducting a survey of current practices and standards in scientific publishing, focusing particularly in the geosciences area, 2) conducting a market validation process where we analyze the needs of the geoscience research community (as well as the more general user community) to develop a clear set of desired features and priorities, and 3) developing a detailed technical design for the software platform. Our proposal also considers how the technology will be supported in the long term. To address this, our transition plan has two complementary parts.  We plan to develop open-source technology that can be used by the academic publishing community, but that can also support a commercial Software-as-a-Service (SAAS) product for evidence-based collaboration for enterprise customers.